U.S. patent number 6,735,515 [Application Number 10/166,980] was granted by the patent office on 2004-05-11 for method and system for providing an electronic horizon in an advanced driver assistance system architecture.
This patent grant is currently assigned to Navigation Technologies Corp.. Invention is credited to Stephan Bechtolsheim, Larry Dunn, Jerry Feigen, Andreas Hecht, Michele Roser, Matthias Schmitt.
United States Patent |
6,735,515 |
Bechtolsheim , et
al. |
May 11, 2004 |
Method and system for providing an electronic horizon in an
advanced driver assistance system architecture
Abstract
A method and system for use by driver assistance systems
installed in a motor vehicle to continuously provide such systems
with updated data about paths along roads onto which the motor
vehicle can travel from a current position of the motor vehicle as
the motor vehicle travels along said roads. The method includes
accessing a database that contains data that represents segments of
roads and intersections of a road network located in a geographic
region in which the motor vehicle is traveling and determining one
or more paths along roads onto which the motor vehicle can travel
from a current position of the motor vehicle. Each path is extended
out to a threshold. Data representing each of the paths is provided
in an organized data structure for use by the driver assistance
systems.
Inventors: |
Bechtolsheim; Stephan (Buffalo
Grove, IL), Dunn; Larry (Lacrosse, IN), Hecht;
Andreas (Frankfurt, DE), Schmitt; Matthias
(Hessen, DE), Feigen; Jerry (Chicago, IL), Roser;
Michele (Chicago, IL) |
Assignee: |
Navigation Technologies Corp.
(Chicago, IL)
|
Family
ID: |
23856381 |
Appl.
No.: |
10/166,980 |
Filed: |
June 10, 2002 |
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
467608 |
Dec 20, 1999 |
6405128 |
|
|
|
Current U.S.
Class: |
701/532; 340/990;
701/454; 701/461 |
Current CPC
Class: |
G01C
21/3667 (20130101) |
Current International
Class: |
G01C
21/34 (20060101); G01C 021/00 () |
Field of
Search: |
;701/208,200 ;73/178R
;340/988,990 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Cuchlinski, Jr.; William A.
Assistant Examiner: Hernandez; Olga
Attorney, Agent or Firm: Kozak; Frank J. Shutter; Jon D.
Kaplan; Lawrence M.
Parent Case Text
REFERENCE TO RELATED APPLICATION
The present application is a continuation of Ser. No. 09/467,608,
filed Dec. 20, 1999, now U.S. Pat. No. 6,405,128, the entire
disclosure of which is incorporated herein by reference.
Claims
We claim:
1. A data architecture for a motor vehicle for providing
continuously updated data about paths along roads onto which the
motor vehicle can travel from a current position of the motor
vehicle as the motor vehicle travels along said roads, the data
architecture comprising: a map database containing data about roads
in a geographic region; a vehicle positioning program that uses
data from sensors to provide an output indicating a current
location along a road segment represented by data in said map
database; a data horizon program that uses the output of the
vehicle positioning program and data from the map database to
determine a plurality of possible paths for the motor vehicle to
travel extending from said current location to an extent; and a
data repository for storing an electronic horizon data object
comprising data representing the possible paths determined by the
data horizon program.
2. A data architecture for a motor vehicle for providing
continuously updated data about paths along roads onto which the
motor vehicle can travel from a current position of the motor
vehicle as the motor vehicle travels along said roads, the data
architecture comprising: a map database containing data about roads
in a geographic region; a vehicle positioning program that uses
data from sensors to provide an output indicating a current
location along a road segment represented by data in said map
database; a data horizon program that uses the output of the
vehicle positioning program and data from the map database to
determine one or more paths that the motor vehicle can travel
extending from said current location to an extent; and a data
repository for storing data representing the possible paths
determined by the data horizon program, wherein said data stored in
said data repository comprises: an electronic horizon data object
containing data attributes representing road segments that form the
paths; and a plurality of electronic horizon objects, wherein each
electronic horizon object includes: a reference to said electronic
horizon data object, and data indicating a distance by which a
vehicle position associated with the represented electronic horizon
object is displaced from the vehicle position associated with the
referenced electronic horizon data object.
3. The invention of claim 2 wherein said data attributes
representing road segments contained in said electronic horizon
data object include data indicating road geometry, data indicating
road attributes, and data representing road objects.
4. The invention of claim 1 further comprising: a data distributor
responsive to the storing of data representing the paths in said
data repository, wherein said data distributor sends messages
indicating availability of new data each time new data is stored in
said data repository.
5. The invention of claim 4 further comprising: a listener program,
wherein said listener program receives said messages from said data
distributor.
6. The invention of claim 5 wherein said listener program is
associated with a driver assistance application that uses the data
stored in said data repository representing the paths.
7. The invention of claim 5 wherein said listener program provides
an output to an in-vehicle bus to which is connected a driver
assistance application that uses the data stored in said data
repository.
8. The invention of claim 7 wherein said in-vehicle bus comprises a
CAN bus.
9. The invention of claim 5 wherein said listener program includes
a queue containing said messages received most recently
thereby.
10. The invention of claim 4 wherein said data distributor sends
messages each time a new vehicle position is determined.
11. The invention of claim 4 further comprising: a listener program
that registers with the data distributor to receive said messages
therefrom.
12. The invention of claim 11 wherein said listener program
receives messages about only one type of data.
13. The invention of claim 11 wherein said listener program
receives messages about more than one type of data.
14. The invention of claim 1 further comprising: a distributor
responsive to the storing of data representing the paths in said
data repository, wherein said data distributor sends said data each
time new data is stored in said data repository.
15. The invention of claim 1 further comprising: a data distributor
responsive to the storing of data representing the paths in said
data repository, wherein said data distributor sends said data via
at least one of point-to-point transmission, multicast
transmission, and broadcast transmission.
16. The invention of claim 1 wherein said data repository also
stores data representing previous locations of the motor
vehicle.
17. The invention of claim 16 further comprising: a data
distributor responsive to the storing of data representing the
paths in said data repository and data representing a new vehicle
location, wherein said data distributor sends messages indicating
availability of new data each time new data is stored in said data
repository.
18. The invention of claim 1 wherein said data repository also
stores sensor data.
19. The invention of claim 18 further comprising: a distributor
responsive to the storing of data representing the paths and said
sensor data in said data repository, wherein said data distributor
sends messages indicating availability of new data each time new
data is stored in said data repository.
20. The invention of claim 1 wherein said map database is located
in said motor vehicle.
21. The invention of claim 1 wherein said data repository also
stores data representing a primary path.
22. The invention of claim 21 wherein said primary path is a
route-based path.
23. The invention of claim 21 wherein said primary path is a
local-road-network-based most likely path.
24. The invention of claim 1 further comprising: a path evaluator
program indicating a most likely path based only on the local road
network.
25. The invention of claim 1 wherein said data about roads
comprises data about road objects including road signs and
crosswalks.
26. The invention of claim 1 further comprising: a routine in said
data horizon program that calculates road curvature using
coordinates of points along roads.
27. A method of providing data representing paths that a vehicle
can take along roads from a current position of the vehicle at a
position along a road, the method comprising: determining a vehicle
position; determining all the available paths along road segments
for the vehicle to travel from said current position out to an
extent associated with a threshold, said available paths include
paths leading to a destination of the motor vehicle and paths not
leading to the destination; storing data representing said
available paths; and making said stored data accessible to
applications that use the data to provide assistance to a driver of
said vehicle while driving.
28. A method of providing data representing paths that a vehicle
can take along roads from a current position of the vehicle at a
position along a road, the method comprising: determining a vehicle
position; determining all the paths along road segments that the
vehicle can travel from said current position out to an extent
associated with a threshold; storing data representing said paths;
and making said stored data accessible to applications that use the
data to provide assistance to a driver of said vehicle while
driving, wherein said storing step comprises: storing a first
object that includes data attributes representing said paths; and
storing a second object that includes a distance by which a vehicle
position associated with said second object differs from a vehicle
position associated with said first object.
29. The method of claim 27 further comprising: after the step of
storing data representing said paths, providing a notification to
said applications that use the data that new data is available.
30. The method of claim 29 further comprising: after the step of
providing a notification, sending the data representing said paths
to an application that responds to said notification.
31. The method of claim 27 further comprising: making stored data
associated with several most recently determined vehicle positions
available to said applications that use the data.
32. The method of claim 27 wherein said stored data is accessible
to applications that use the data over a CAN bus.
33. A data architecture for a motor vehicle for providing
continuously updated data about paths along roads onto which the
motor vehicle can travel from a current position of the motor
vehicle as the motor vehicle travels along said roads, the data
architecture comprising: a data horizon program that temporarily
stores in a data repository data representing road segments located
around the motor vehicle; and a data listener associated with an
application that provides driver assistance features, wherein said
data listener receives notifications from said data horizon program
about newly stored data and obtains said data from said data
horizon program, as needed.
34. A data architecture for a motor vehicle for providing
continuously updated data about paths along roads onto which the
motor vehicle can travel from a current position of the motor
vehicle as the motor vehicle travels along said roads, the data
architecture comprising: a map database containing data about roads
in a geographic region; a vehicle positioning program that uses
data from sensors to provide an output indicating a current
location along a road segment represented by data in said map
database; a data horizon program that uses the output of the
vehicle positioning program and data from the map database to
determine a plurality of possible paths for the motor vehicle to
travel extending from said current location to an extent, said
plurality of possible paths include paths leading to a destination
of the motor vehicle and paths not leading to the destination; and
a data repository for storing data representing the possible paths
determined by the data horizon program.
Description
BACKGROUND OF THE INVENTION
The present invention relates to a map data architecture platform
that can be used in on-road vehicles, such as automobiles, trucks,
buses, and so on, and in particular the present invention relates
to a map data architecture platform that supports advanced driver
assistance systems provided in on-road vehicles.
Advanced driver assistance systems ("ADAS") have been developed
with the intent of improving the safety, comfort, efficiency, and
overall satisfaction of driving. Examples of advanced driver
assistance systems include adaptive headlight aiming, adaptive
cruise control, and adaptive shift control. Adaptive headlight
aiming adjusts the vehicle's headlights, i.e., width, rotational
angle, elevation angle, and brightness, based on the curvature of
the road ahead of the vehicle, tilt, elevation change, and other
factors. Adaptive cruise control maintains and/or resumes a set
speed or safe following distance from other vehicles at slower than
the set speed based on data about vehicle speed, nearby vehicles
and other obstructions, type of road traveled (motorway vs. local
road), road curvature, tilt, elevation, and other factors. Adaptive
shift control adjusts the gearing and shifting of automatic
transmissions based on sensor data about vehicle speed, engine
speed, road curvature, tilt, elevation, and other factors. There
are other advanced driver assistance systems in addition to
these.
These advanced driver assistance systems use a variety of sensor
mechanisms in the vehicle to determine the current state of the
vehicle and the current state of the roadway in front of the
vehicle. These sensor mechanisms may include radar and
vision-oriented sensors, such as cameras. Although radar and
vision-oriented sensors are important components of advanced driver
assistance systems, these components have limitations. The range
and/or accuracy of radar or vision-oriented sensors can be affected
by certain environmental conditions, such as fog, heavy rain or
snow, or snow-covered roads. Moreover, radar and vision-oriented
systems do not reliably detect certain useful road attributes, such
as speed limits, traffic signs, bridge crossings, etc. Further,
radar and vision-oriented sensors cannot "see" around corners or
other obstructions and therefore may be limited under such
circumstances.
One way to address the limitations of radar and vision-oriented
systems is to use digital map data as an additional component in
advanced driver assistance systems. Digital map data can be used in
advanced driver assistance systems to provide information about the
road ahead. Digital map data are not affected by environmental
conditions, such as fog, rain or snow. In addition, digital map
data can provide useful information that cannot reliably be
provided by vision-oriented systems, such as speed limits, traffic
and lane restrictions, etc. Further, digital map data can be used
to determine the road ahead of the vehicle even around corners or
beyond obstructions. Accordingly, digital map data can be a useful
addition in advanced driver assistance systems.
Although digital map data can be used as an additional component in
advanced driver assistance systems, issues remain to be addressed
before digital map data can be widely used for such purposes. For
example, there is a need to efficiently handle the relatively large
amount of digital map data required for advanced driver assistance
systems. In addition, different advanced driver assistance systems
require different types and quantities of digital map data and
therefore there is a need to provide those digital map data needed
by the various advanced driver assistance systems.
SUMMARY OF THE INVENTION
To address these and other objectives, the present invention
comprises a method and system for use by driver assistance systems
installed in a motor vehicle to continuously provide such systems
with updated data about paths along roads onto which the motor
vehicle can travel from a current position of the motor vehicle as
the motor vehicle travels along said roads. The method includes
accessing a database that contains data that represents segments of
roads and intersections of a road network located in a geographic
region in which the motor vehicle is traveling and determining one
or more paths along roads onto which the motor vehicle can travel
from a current position of the motor vehicle. Each path is extended
out to a threshold. Data representing these paths are provided in
an organized data structure for use by the driver assistance
systems. The data representing the paths include data about road
geometry, road attributes, and objects along each path.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a functional block diagram of the advanced driver
assistance systems map data architecture 100.
FIG. 2 is a block diagram of the sensor component of the advanced
driver assistance systems map data architecture 100 shown in FIG.
1.
FIGS. 3A and 3B show the types of data contained in the map
database component of the advanced driver assistance systems map
data architecture.
FIG. 4 is a block diagram of the components of the software tools
shown in FIG. 1.
FIG. 5 is a block diagram of the components of the data engine
shown in FIG. 1.
FIG. 6 is an illustration of a portion of a road network with a
depiction of an electronic horizon superimposed thereupon.
FIG. 7 is an illustration of segments identifiers used when
describing paths in an electronic horizon.
FIG. 8 is an illustration of path descriptors in an electronic
horizon.
FIG. 9 is an illustration of a path descriptor for a U-turn.
FIG. 10 is a block diagram showing components of the data
repository in FIG. 1.
FIG. 11 is a block diagram showing components used for the storage
of electronic horizon data in the data repository in FIG. 10.
FIG. 12 is a block diagram showing additional components of the
data repository in FIG. 10.
FIG. 13 is a block diagram showing components of the data
distributor in FIG. 1.
FIG. 14 is a block diagram showing components of the data listener
in FIG. 1.
FIG. 15 is a block diagram showing an embodiment of a driver
assistance application associated with plural listeners.
FIG. 16 is a block diagram showing how a driver assistance
application uses various functions to obtain electronic horizon
data.
FIG. 17 is a block diagram showing an alternative embodiment of the
driver assistance systems map data architecture.
DETAILED DESCRIPTION OF THE DESCRIPTION OF THE PRESENTLY PREFERRED
EMBODIMENTS
I. Terminology
The following terminology and concepts are used in this
specification. (The terminology and definitions provided herein are
not intended to be limiting. Other terminology and definitions may
be used to express similar or identical concepts.)
(1) Segments and nodes. A "segment" (also referred to as a "road
segment") is a length of a road. Each segment has two end points. A
"node" is one of the end points of a segment. A segment has a left
node and a right node. The left node is the node with the smaller
longitude value. If the longitude values of both nodes are the
same, the left node is the node with the smaller latitude.
According to one embodiment, segments and nodes are represented by
data in a map database used by the driver assistance system map
data architecture.
(2) Shape points. "Shape points" are intermediate points on a
segment between its end points. Shape points are used for several
purposes. Shape points may be used to model the curvature of a road
segment. Shape points may also be used to model overpasses and
underpasses. For example, when one road segment crosses another
road segment at a different elevation (e.g., an overpass or
underpass), a shape point is associated with each road segment at
the location of the crossing and an attribute of each shape point
indicates a relative altitude or an absolute altitude of the
associated road segment at that location. According to one
embodiment, shape points are represented in the map database used
by the driver assistance system map data architecture.
(3) Travel direction. The "travel direction" on a segment (the
permissible direction of a vehicle travel on a segment) is
expressed in terms of "travel from the left node to the right node"
or "from the right to the left node."
(4) Entrance node and exit node. The node encountered first in the
context of travelling on a segment is referred to as the "entrance
node," the other node is referred to as the "exit node."
(5) Point. A "point" refers to a node or a shape point of a
segment. A point has a geographic location (e.g., latitude,
longitude, and altitude) associated with it.
(6) Segment location. A "segment location" is any place on a
segment. Whereas the term "point" only refers to nodes and shape
points of segments, a segment location includes all locations on a
segment including the nodes, all shape points, and all logical
points (i.e., locations) between the nodes and shape points.
(7) Segment bearings and headings. The "bearing" of a segment at a
node refers to the direction of the segment at that node. The
direction is measured from the node towards the inside of the
segment. For instance, the bearing at the left node is the heading
of a vehicle at the left node as the vehicle travels from the left
to the right node. The heading of a segment at the left or right
node is computed from the bearing value at the appropriate node
plus 180 degrees.
(8) Curvature. "Curvature" describes how a portion of a segment
curves at a point or a segment location. There are different ways
of calculating and representing curvature. One way to represent the
curvature at a point of a segment is by the radius of a circle that
corresponds to the curve of the segment at that point. According to
one embodiment, curvature is represented by data in a map database
used by the driver assistance system map data architecture.
According to another embodiment, curvature may be calculated using
data indicating the coordinates of successive points along a road
segment.
(9) Path. A "path" is a sequence of one or more road segments (or
portions thereof) upon which a vehicle might travel from a current
location.
(10) Road objects. A "road object" refers to an object located on
or along a road, such as a sign or a crosswalk.
(11) Road geometry. "Road geometry" refers to the shape and
curvature of a road. Road shape is defined by the geographic
coordinates of points along a road segment. ("Curvature" is
described separately below.)
II. Advanced Driver Assistance Systems Map Data Architecture
A. Overview
FIG. 1 is a functional block diagram of the advanced driver
assistance systems map data architecture 100. The advanced driver
assistance systems map data architecture 100 is a combination of
software and hardware components installed in a motor vehicle 108.
The advanced driver assistance systems map data architecture 100
provides access to map-related data for use by advanced driver
system applications 200. The advanced driver assistance systems map
data architecture 100 includes the following components.
(1). Sensors 120.--The sensors 120 provide outputs that are used by
programming in the advanced driver assistance systems map data
architecture 100 to determine the position of the vehicle 108 on
the road network and to provide other information, such as speed
and heading of the vehicle. (In addition to these sensors 120, the
advanced driver system applications 200 may use the outputs from
other types of sensors 122. These other types of sensors 122 may
include radar or vision system-types of sensors.)
(2). A map database 130.--The map database 130 includes information
about geographic features, such as roads and points of interest, in
the geographic area in which the vehicle 108 in which the advanced
driver assistance systems map data architecture 100 is installed is
traveling.
(3). Data horizon program 110.--The driver assistance systems map
data architecture 100 includes a data horizon program 110. The data
horizon program 110 includes the programming that uses the map
database 130 and inputs from the sensors 120 to provide map-related
data to the advanced driver assistance systems 200.
(4). Software tool components 150.--In this embodiment, the
software tool components 150 are a part of the data horizon program
110. The software tool components 150 include programming for
accessing the map database 130 and software tool programs for
performing certain required functions with the map data obtained
from the map database 130.
(5). A monitoring program 160.--The monitoring program 160 is a
software component of the advanced driver assistance systems map
data architecture 100 that provides for monitoring execution of the
data horizon program 110.
(6). A configuration program 165.--The configuration program 165 is
a software component of the advanced driver assistance systems map
data architecture 100 that provides for configuration of the data
horizon program 110.
(7). A data engine 170.--The data engine 170 is a component of the
data horizon program 110. The data engine 170 determines and
obtains from the map database 130 the relevant data about the road
lying ahead of (or behind) the vehicle.
(8). A data repository 180.--The data repository 180 is a component
of the data horizon program 110. The data repository 180 contains
the latest relevant data about the road lying ahead of (or behind)
the vehicle as determined by the data engine 170.
(9). A data distributor 190.--The data distributor 190 is a
component of the data horizon program 110. The data distributor 190
provides notification that new data about the road lying ahead of
(or behind) the vehicle has been stored in the data repository
180.
(10). One or more advanced driver assistance applications
200.--These applications 200 use the map-related data provided by
the data horizon program 110. These applications 200 may include
adaptive headlight aiming, adaptive cruise control, obstruction
detection, obstruction avoidance, collision avoidance, adaptive
shift control and others.
(11). One or more data listeners 300.--A data listener 300 is a
software component used for obtaining data from the data horizon
program 110. A data listener 300 receives the notifications from
the data distributor 190 and obtains data from the data repository
180. A data listener 300 may be implemented as part of each driver
assistance application 200 or a data listener may be implemented as
a standalone software component.
Each of the above components of the advanced driver assistance
system map data architecture 100 is described in more detail
below.
B. The positioning sensors 120
Referring to FIGS. 1 and 2, the data horizon program 110 receives
the outputs from the positioning sensors 120. According to one
embodiment, these sensors 120 include a GPS system 120(1), a
gyroscope 120(2), a vehicle speed sensor 120(3) and a vehicle
forward/reverse sensor 120(4). Other types of sensors 120(5) may
also be included. For example, the sensors may include inertial
navigation sensors.
In one embodiment, the GPS system 120(1) is a system manufactured
by Trimble and the gyroscope 120(2) is a unit manufactured by
Murata. Equipment from other manufacturers may also be used. Data
indicating the vehicle speed and/or the vehicle forward/reverse
direction may be obtained from sensors or components provided for
other purposes elsewhere in the vehicle 108. In one embodiment, the
gyroscope and speed signals are collected every 100 milliseconds.
The GPS and the vehicle forward/reverse direction sensor are
provided at a frequency of once per second. In other embodiments,
the sensor outputs may be provided at different frequencies.
C. The map database(s) 130
(1). Map database organization
Referring again to FIG. 1, the map database 130 includes
information about roads, intersections, points of interest, and
possibly other geographic features in the geographic region in
which the vehicle 108 in which the advanced driver assistance
systems map data architecture 100 is installed is traveling. In the
embodiment shown in FIG. 1, the map database 130 is formed of one
or more component databases. Specifically, the map database 130
includes a primary database 130(1) and a supplementary database
130(2).
The primary map database 130(1) may be similar or identical to a
database used in in-vehicle navigation systems. According to this
embodiment, the primary map database 130(1) supports
navigation-related functions, including vehicle positioning, route
calculation, route guidance, and map display. The primary database
130(1) also provides support for a portion of the advanced driver
assistance system functions. In this embodiment, the primary
database 130(1) also provides a portion of the data readings
provided to the driver assistance applications 200, as described
below.
In one embodiment, the primary map database 130(1) is a database in
the SDAL.TM. physical storage format developed and published by
Navigation Technologies Corporation of Rosemont, Ill. In one
embodiment, the primary database 130(1) is in version 1.7 of the
SDAL.TM. physical storage format. A suitable embodiment of a
primary map database is described in U.S. Pat. No. 5,968,109, the
entire disclosure of which is incorporated by reference herein.
(The inventive subject matter disclosed herein is not limited to
any specific database format.)
The supplementary database 130(2) also contains data about roads
and intersections in the geographic region. However, the
supplementary database 130(2) includes data that is not necessarily
provided in the primary map database 130(1). The supplementary map
database 130(2) may include higher quality (i.e., more accurate)
data than the data which is contained is in the primary database
130(1). For example, with respect to road geometry, the data in the
supplementary database 130(2) may be more accurate with respect to
longitude, latitude, and/or altitude. Also, the starting and
stopping locations of tunnels may be more accurately specified in
the supplementary database 130(2) than in the primary database
130(1). Further, the data in the supplementary database 130(2)
maybe more accurate with respect to derived information, such as
curvature.
The supplementary database 130(2) may also include more kinds of
data (e.g., more kinds of attributes) than the data which is
contained in the primary database 130(1). For example, the
supplementary database 130(2) may include data about road objects,
such as signs and crosswalks including their positions along the
road segment, sign object type and sign text. The supplementary
database 130(2) may also include data indicating whether a road is
tree-lined, etc.
According to one embodiment, both the primary database 130(1) and
the supplementary database 130(2) include data representing all the
roads and intersections in the covered region. According to this
alternative, the data in the supplementary database 130(2)
supplements the representation of each road segment which is also
represented in the primary database 130(1).
According to an alternative embodiment, the supplementary database
130(2) represents fewer roads than the primary database 130(1). In
this alternative embodiment, whereas the primary database 130(1)
may include data representing all the roads and intersections in
the covered region, the supplementary database 130(2) includes data
representing only a portion of all the roads in the covered region.
For example, the supplementary database 130(2) may include only the
roads with the highest traffic volumes (e.g., expressways, major
thoroughfares). The road segments represented by the supplementary
database 130(2) may also be represented by data in the primary
database 130(1).
According to yet another alternative embodiment, instead of using
two separate databases, a single database is used by the driver
assistance systems map data architecture 100. In this single
database embodiment, the lower accuracy data contained in the
primary database 130(1) is combined with the higher accuracy data
contained in the supplementary database 130(2). In the single
database embodiment, all the roads may be represented by data
having the high accuracy standard of the supplementary database.
Alternatively, in the single database embodiment, only some of the
represented roads are represented by higher accuracy data and the
remainder of the roads are represented by a lower accuracy
data.
In the single database embodiment that contains both higher
accuracy data and lower accuracy data, a means is provided to
indicate whether a represented road segment is represented by
higher accuracy data or by lower accuracy data. A data attribute
(e.g., a high-accuracy-data bit) may be associated with each data
entity representing a road segment to indicate whether the data
representing the segments conforms to a specified high accuracy
standard. In further alternatives, roads may be represented by data
of different accuracy levels. Each of these different accuracy
levels may be indicated by a accuracy level designation (e.g.,
0-7).
(2). Integrating data of different accuracy levels
As stated above, in some embodiments of the map database 130 some
roads are represented by data having an accuracy level high enough
for use by advanced driver assistance system applications and other
roads are represented by data having an accuracy level that is not
high enough to be used by advanced driver assistance system
applications. In these embodiments, a means is provided by which
higher accuracy data is integrated with lower (or unknown) accuracy
data. To provide this integration, data are included in the map
database 130 to represent transition segments. A "transition
segment" is a segment which is connected at one end to another
segment represented by data having a high accuracy level and at its
other end to another segment represented by data of a lower (or
unknown) accuracy level. In a transition segment, the coordinates
of the node connected to the segment represented by data having a
high accuracy level are stored at the higher accuracy level.
However, the coordinates of the node connected to the segment
represented by data of a lower (or unknown) accuracy level are
stored to the lower accuracy level. Therefore, according to this
embodiment, there are three classes of segments: (1) segments
represented by high accuracy data, (2) segments represented by
lower (or unknown) accuracy data, and (3) transition segments
connecting (1) and (2).
(3). Kinds of data attributes included in the map database
As stated above, the map database 130 includes information about
roads and intersections. According to one embodiment, the map
database 130 represents each road segment with a separate segment
data entity. Each node at the end point of a road segment is
represented by a separate node data entity. The map database 130
includes (data) attributes associated with the segment data
entities and (data) attributes associated with the node data
entities. Node attributes relate to a property or characteristic of
the end nodes of a segment. Segment attributes relate to a property
or characteristic associated with the segment as a whole or with a
specific point (location) along the segment.
Examples of node attributes include the following:
(1). The number of segments extending from the current node. This
count includes the current segment (i.e., the entrance segment).
All segments are counted, whether accessible or not.
(2). The number of possible turns the vehicle can perform at the
specified node. (U-turns are not included in this count.)
In addition to the above, various other attributes may be
associated with nodes, including geographic coordinates, altitude,
name, identification (e.g., by ID number) of road segments
connected thereto, turn restrictions, etc.
As stated above, segment attributes can relate to a property or
characteristic associated with a specific point (location) along
the segment. These attributes of a segment are referred to as
"point dependent attributes." This type of segment attribute
describes a property related to a "point" on a segment (where a
"point" refers to either one of the two end nodes of a segment or a
shape point on the segment).
A point dependent attribute is used to represent a property of the
road segment at the point, as grade, banking, etc.
In a present embodiment, certain point dependent attributes are
associated with a travel direction along a segment. A "stop sign"
attribute is an example of a point dependent attribute associated
with a travel direction. A "stop sign" attribute indicates the
presence of a stop sign at a point along a segment associated with
a specific direction of travel (e.g., there may not be a stop sign
when traveling in the opposite direction along the segment).
A "stop sign" attribute is also an example of a Boolean attribute.
A Boolean attribute is a point dependent attribute that is either
true or false at the specific point. A "stop sign" is modeled using
a Boolean attribute because a stop sign is either present or not
present at a specific point.
Another kind of point dependent attribute is a Boolean transition
attribute. Boolean transition attributes describe properties or
characteristics that apply to every location on a segment, not just
to the points of a segment. A Boolean transition attribute is an
attribute that changes values only at segment points, if at all.
(The terms "before" and "after" refer to a vehicle approaching a
point and driving beyond that point.) For example, for any location
on a segment (not just any point) given a travel direction, vehicle
"passing" is either allowed or not allowed. In order to model
whether vehicle "passing" is allowed, an assumption is made that
any related sign (such as "start no pass zone" and "end no pass
zone") is located at a point of a segment. If this is the case, one
of the following apply for any point of the segment.
True.fwdarw.true transition: passing is allowed before the point
and after the point.
True.fwdarw.false transition: passing is allowed before the point
but not after the point.
False.fwdarw.true transition: passing is not allowed before the
point but is allowed after the point.
False.fwdarw.false transition: passing is not allowed before the
point and is also not allowed after the point.
Another kind of point dependent attribute is an integer range
transition attribute. An integer range transition attributes is an
attribute representing integer ranges or integer intervals. An
integer range attribute has a fixed value between any two
consecutive points of a segment but may change its value at any
point to a different interval. A value is defined before a point
and at the point. An example of when an integer range transition
attribute is used is for "maximum speed information." An integer
range value such as {20 . . . 29} means that the maximum legal
speed is between 20 and 29 km/h. An integer range value such as {20
. . . 20} means that the maximum legal speed is exactly 20 km/h.
Values may also be specified for different times of day.
Some segment attributes may be identical for every point of the
segment (e.g., a road name). Such attributes may be specified once
for the entire segment.
Point dependent attribute information may be stored in SDAL.TM.
format or in other database tables.
FIGS. 3A and 3B show some of the types of data included in the map
database 130. The table column labeled "Source" in FIGS. 3A and 3B
indicates whether the data element is found in the primary database
130(1), the supplementary database 130(2), or both.
(4). Curvature
According to one embodiment, included among the segment attributes
is an attribute representing curvature. "Curvature" is a property
of a point along a length of a segment. Curvature describes how a
portion of a segment curves at that point. In a present embodiment,
curvature is defined for the points of a segment (i.e., shape
points, nodes). Curvature is described by two components: a
curvature direction (left curve, right curve and straight) and a
curvature radius. No curvature radius is defined for the case of a
straight or nearly straight line. (A segment for which the
curvature radius exceeds a configurable threshold value may be
considered a straight line.)
Curvature data may be obtained in several different ways. One way
to obtain curvature data is to measure it directly using sensor
equipment (e.g., an accelerometer) and storing the measurement as a
data attribute associated with a point in the map database 130.
Another way to obtain curvature data is to compute the curvature
using position data. For a sequence of three points, the curvature
at the middle point can be determined by computing the radius of a
circle whose circumference includes the positions of the three
points. Curvature data obtained by calculation using position data
may be stored in the map database 130. Alternatively, curvature may
be computed as needed by a software function in the vehicle. Such a
software function may be included among the advanced driver
assistance system applications 200 using position data associated
with points stored in the map database 130. Alternatively, a
software function that computes curvature from position data may be
included in the data horizon program 110.
D. Software tool components 150
The software tool components 150 provide the foundation upon which
the data horizon program 110 is built. In the embodiment shown in
FIGS. 1 and 4, the software tool components 150 include a data
access layer 150(1), navigation applications 150(2), and an object
framework 150(3).
The data access layer 150(1) provides for accessing the map
database 130. In one embodiment, the data access layer 150(1) is
the SDAL.TM. library available from Navigation Technologies
Corporation of Rosemont, Ill. The data access layer 150(1) provides
a set of application programming interfaces (API) in the form of
software libraries for efficient access to the map attributes in
the primary database 130(1). An embodiment of the data access layer
150(1) is described in copending application Ser. No. 08/740,298,
filed Oct. 25, 1996, the entire disclosure of which is incorporated
by reference herein.
The navigation applications 150(2) provide functions similar to
those used in in-vehicle navigation systems. According to one
embodiment, the navigation applications 150(2) are provided in the
form of API software library routines. These API software library
routines provide for operations frequently used in map-data-related
applications. Included among the navigation applications 150(2) are
vehicle positioning 150(2)(1), map display 150(2)(2), route
calculation 150(2)(3), geo-coding 150(2)(4), and direction guidance
150(2)(5). In one embodiment, the navigation applications 150(2)
are NavTools.TM. software available from Navigation Technologies
Corporation of Rosemont, Ill. Embodiments of navigation
applications for vehicle positioning, map display, route
calculation and direction guidance are described in copending
applications Ser. Nos. 09/276,377, 09/047,141, 09/047,698,
08/893,201, and 09/196,279, the entire disclosures of which are
incorporated by reference herein.
The object framework 150(3) provides an object-oriented wrapper
around the data access layer 150(1) and the navigation applications
150(2). The object framework 150(3) simplifies use of the data
access layer 150(1) and the navigation applications 150(2). The
object framework 150(3) may also facilitate development of
applications on certain platforms (e.g., a Microsoft Windows/NT
environment).
XML in the advanced driver assistance system data architecture
In one embodiment, the advanced driver assistance system map data
architecture 100 uses XML (extensible Markup Language). For
example, log file and other information may be generated in XML.
Similarly, some of the information read in by the advanced driver
assistance system map data architecture 100 may be encoded in XML.
An advantage of having one file format for multiple purposes
simplifies the manipulation and further processing of input and
output information. Use of XML is advantageous in a development and
testing environment.
In one embodiment, Microsoft's Internet Explorer Version 5.0 (IE5)
or another program that supports XML as a native file format can be
used. IE5 also processes XSL (XML-related style files). This allows
XML files to be presented in different ways.
E. The data engine 170
(1). Overview
Referring to FIGS. 1 and 5, the data horizon program 110 includes a
data engine 170. The data engine 170 is that component of the data
horizon program 110 that calculates an electronic horizon
(described in more detail below). The data engine 170 provides an
output that includes the data representing the electronic horizon
in an organized format. The data engine 170 provides this output on
a cyclic basis.
(2). Inputs to the data engine
In performing its functions, the data engine 170 uses data
indicating the vehicle position (including direction and speed) as
an input. Referring to FIG. 5, the data engine 170 includes a data
receiving process 170(1) that performs this function. The data
receiving process 170(1) receives data indicating the vehicle
position from the vehicle positioning tool 150(2)(1). The data
indicating the vehicle position includes an identification of the
road segment upon which the vehicle is located, the position along
the identified road segment at which the vehicle is located, and
the direction the vehicle is heading along the road segment. The
road segment upon which the vehicle is located is determined by the
vehicle positioning tool 150(2)(1) using data from the map database
130.
The position along the identified road segment may be provided in
various different ways. For example, the vehicle position along the
road segment may be provided as a distance from an end (e.g., n
meters from the left endpoint). In another alternative, if the road
segment includes shape points located between its end points, the
position of the vehicle along the road segment may be indicated by
that shape point to which the vehicle is closest. Alternatively,
the vehicle position along the identified road segment may be
identified as that shape point which is immediately ahead of the
vehicle position. In another alternative, the vehicle position
along a road segment may be provided in incremental portions of the
road segment length (e.g., n/256.sup.th along a road segment).
The data indicating the vehicle direction may be provided to the
data receiving component 170(1) of the data engine 170 by the
vehicle positioning tool 150(2)(1) by indicating which node of the
segment the vehicle is heading toward. The data receiving component
170(1) of the data engine 170 also obtains the speed of the vehicle
(e.g., from the sensors 120).
The vehicle positioning tool 150(2)(1) may provide a new output
indicating a new vehicle position at regular intervals. These
intervals may be once per second, 10 times per second, 100 times
per second, once every 2 seconds, or any other period. The
intervals may also be irregular intervals or may be intervals based
on some other factor, such as distance, or a combination of
factors, such as time and distance. According to a present
embodiment, the data receiving component 170(1) receives each
output of the vehicle positioning tool 150(2)(1) indicating a new
vehicle position.
The vehicle positioning tool 150(2)(1) may determine that the
vehicle 108 is off-road. The vehicle 108 is off-road if the vehicle
positioning tool 150(2)(1) cannot determine a position of the
vehicle along a road segment represented in the map database 130.
This may occur if the vehicle is actually off-road (e.g., not on
any road segment, such as in a parking lot, in a field, or outside
the coverage region of the map database 130). Alternatively, the
vehicle positioning tool 150(2)(1) may determine that the vehicle
is off-road if reliable sensor information cannot be obtained. If
the vehicle positioning tool 150(2)(1) indicates that the vehicle
is off-road, information indicating this off-road status is
provided to the data receiving process 170(1). The determination of
an electronic horizon requires a valid vehicle position with the
vehicle positioned on a specific location of a specific segment. If
the vehicle is off-road, the data engine 170 does not calculate an
electronic horizon.
(3). Calculation of the electronic horizon
The data engine 170 includes an electronic horizon calculation
process 170(3). The electronic horizon calculation process 170(3)
determines which road segments and intersections should be
represented in the output of the data engine 170. These segments
and intersections represented in the output of the data engine 170
are the potential paths the vehicle may follow from the current
vehicle position. The extent that each of these potential paths
extends from the current vehicle position is determined by the
electronic horizon calculation process 170(3). The "electronic
horizon" refers to the collection of the roads and intersections
leading out from the current vehicle position to the extents
determined by the electronic horizon calculation process 170(3).
Thus, the "electronic horizon" represents the road ahead of (or
possibly behind) the vehicle. The electronic horizon is also a
representation of potential driving paths of the vehicle from the
current vehicle position. The "electronic horizon" also refers to
the collection of data that represents the roads and intersections
leading out from the current vehicle position to the aforementioned
extents, including the road attributes, road objects, and road
geometry of the road segments that form the electronic horizon.
To perform the function of determining the electronic horizon, the
electronic horizon calculation process 170(3) obtains the data
indicating the vehicle's current position from the data receiving
process 170(1). Using the data indicating the vehicle's current
position, the electronic horizon calculation process 170(3) obtains
data from the map database 130 that relates to all the roads around
the vehicle's current position. The data engine 170 includes a
component process 170(2) that obtains these data from the map
database 130. If the map database 130 includes both a primary
database and a supplementary database, the component process 170(2)
combines the primary and secondary data for use by the data
engine.
After obtaining data that relate to all the road segments around
the vehicle's current position, the data engine 170 determines
which road segments represent the electronic horizon. This step
includes determining the extents (or boundaries) of the electronic
horizon. In determining the extents of the electronic horizon, the
electronic horizon calculation process 170(3) provides that the
potential paths extending from the current vehicle position are
sufficiently large so that the driver assistance applications 200
(in FIG. 1) that use the data output by the data horizon program
110 are provided with all the data they may need to perform their
functions, given the speed and direction of the vehicle as well as
specific requirements of each of the driver assistance applications
200. On the other hand, the electronic horizon calculation process
170(3) builds an electronic horizon as small as possible in order
to reduce the computational resources required to build it and also
to reduce the computational resources required by the driver
assistance applications 200 when using the data included in the
electronic horizon.
The extents of the electronic horizon are determined using one or
more costing functions, as explained in more detail below. Briefly,
starting with the segment upon which the vehicle is currently
located, each segment of each path leading away from the current
vehicle position is evaluated for possible inclusion in the
electronic horizon. The electronic horizon calculation process
170(3) stops evaluating segments to add to a path from the current
vehicle position when the path has at least a minimum threshold
cost, if possible. The electronic horizon calculation process
170(3) stops calculating an electronic horizon when all segments
included in all the paths from the current vehicle position are
determined. When the electronic horizon calculation process 170(3)
stops calculating an electronic horizon, the extents of the
electronic horizon are determined.
According to one embodiment, the electronic horizon is represented
by a tree from which the potential driving paths from the vehicle's
current location diverge as branches. The electronic horizon
calculation process 170(3) forms this tree when determining which
road segments and intersections to include in the electronic
horizon. The tree that forms the electronic horizon includes
components by which each point along each path can be specified and
defined within the context of the entire tree structure. In this
manner, formation of the electronic horizon is done in a
consistent, reliable and reproducible manner. This provides
features, such as a level of confidence, that can be used by the
advanced driver assistance systems 200.
(4). Electronic horizon terminology
The components of the electronic horizon are organized so that the
driver assistance applications 200 can use the data representing
the roads located around the vehicle. The electronic horizon
components include the following:
(a) First segment. The road segment upon which the vehicle is
located is the "first segment" of the electronic horizon tree.
(b) Root node. The entrance node of the first segment of the
electronic horizon is the "root node" of the electronic horizon
tree.
(c) Internal node. An "internal node" of an electronic horizon is a
node to which at least two segments of the electronic horizon are
attached.
(d) Entrance and exit segments. Each internal node of an electronic
horizon has one "entrance segment", i.e., a segment on which the
vehicle can potentially drive towards that node. An internal node
also has one or more "exit segments", i.e., segments on which a
vehicle potentially drives away from the current internal node.
(e) Leaf node. A "leaf node" is a node within an electronic horizon
where no additional segments are attached.
(f) Accessible sub-tree. Those segments of the electronic horizon
that are accessible by legally permitted paths from the first
segment of the electronic horizon form the "accessible sub-tree" of
the electronic horizon.
(g) Simple path electronic horizon. An electronic horizon is
referred to as a "simple path electronic horizon" if the accessible
sub-tree of it consists of a linear list of segments only.
(h) Single segment electronic horizon. A simple path electronic
horizon is referred to as a "single segment electronic horizon" if
the accessible sub-tree of the electronic horizon consists of a
single segment only.
(i) Inaccessible segment. An "inaccessible segment" is segment
which is connected to a node included in the electronic horizon but
which cannot be entered legally from the node. For example, the
segment may be a one way street and the direction of the one way
restriction is such that it is illegal to drive onto the segment
from the node that is part of the electronic horizon.
Alternatively, there may be a turn restriction in effect that does
not allow a vehicle to turn onto the segment from the node which is
part of the electronic horizon. Note that a particular segment may
be inaccessible if the vehicle approaches the segment via one node
but accessible if the vehicle approaches the segment via a
different node. Formation of the electronic horizon may be
configured (e.g., through the costing function, as described below)
so that inaccessible segments are included in an electronic horizon
or alternatively formation of the electronic horizon may be
configured so that inaccessible segments are excluded from an
electronic horizon.
EXAMPLE
FIG. 6 illustrates an electronic horizon superimposed on a portion
of the road network. In FIG. 6, the inaccessible segment is
excluded from the electronic horizon sub-tree.
(5). Component identification of electronic horizon
The electronic horizon includes a means by which each of the paths
leading from the current vehicle position out to the extents of the
electronic horizon can be uniquely identified. Each of the
component parts of an electronic horizon can be identified using
segment identifiers, path identifiers, segment descriptors, node
descriptors and point descriptors.
(a) Segment identifiers. A "segment identifier" identifies a
segment with an index number with respect to a particular node. The
entrance segment of a node has an index of 0. Exit segments of a
node are indexed starting at 1. All exit segments of a node are
marked clockwise. The first segment (i.e., index=1) is the segment
that follows the entrance segment in a clockwise direction. It is
possible that no exit segment exists for a particular node (e.g., a
leaf node). FIG. 7 illustrates the assignment of segment
identifiers at an intersection (i.e., a node).
(b) Path descriptors. A "path descriptor" describes a path by a
list of the segment identifiers of that path. Because every path
includes the first segment of an electronic horizon, every path
descriptor starts with 0. Any segment after the first electronic
horizon segment is identified by its segment identifier with
respect to its entrance node. FIG. 8 shows an example of how path
descriptors are formed. FIG. 8 shows the same electronic horizon as
shown in FIG. 6. Next to each segment in the electronic horizon is
its segment identifier defined with respect to the entrance node
thereto. FIG. 8 also includes a table of path descriptors for each
of the paths in the electronic horizon.
Note that under some circumstances, a segment contained in an
electronic horizon can be entered by more than one path. If a
segment can be entered by more than one path, the segment is
included in each of the path descriptors. Thus, a segment can be
included more than once in a description of an electronic
horizon.
Sometimes it is necessary to define an invalid path. Such a path
has a path descriptor of -1.
Path descriptors can also be used to describe paths involving
U-turns. FIG. 9 shows an example of how a path descriptor can be
used to describe a U-turn. In FIG. 9, a vehicle traveling from a
current vehicle position to node A, then node B and then back to
node A would travel the path 0.2.0. In any path, segment 0 is the
segment on which the vehicle travels towards a node. Therefore, to
describe a U-turn, segment index 0 is used to indicate that the
vehicle exits the node on the same segment it has driven towards
the node.
(c) Order of Paths. All paths of an electronic horizon define a
complete order. Because the number of paths is finite, a "first
path" and a "last path" of an electronic horizon exist. Given two
paths descriptors, p1 and p2, this order is defined as follows.
Repeatedly compare the individual segment indices of the two path
descriptors. In each iteration, the following steps are
executed:
First, if the first two individual segment indices are identical,
continue by comparing the next pair of segment indices. For
example, assume two path descriptors 0.4.3.1 and 0.4.2.2. The
comparison computation at this point has reached the second segment
index (`4` in both cases).
The two individual segment indices are different. In this case the
path descriptor with the smaller segment index value is deemed to
be smaller than the path descriptor with the larger of the two
values. For example, assume two path descriptors 0.2.3.1 and
0.2.4.2. Comparing the third segment indices `3` and `4` of both
path descriptors leads the first path descriptor to be declared as
being smaller than the second path descriptor. The comparison
operation stops at this point.
The number of segment indices for both paths has been exceeded. The
two path descriptors are identical in this case and the path
comparison computation stops. An example would be two path
descriptors 0.1.2 and the preceding computation has just compared
the third of the segment descriptors (`2`).
The number of segment indices for the first path descriptor has
been exceeded, but there is another segment index still available
for the second path descriptor. In this case, the first path
descriptor is deemed to be smaller than the second path descriptor
and the path comparison operation stops at this point. For
instance, assume path descriptors 0.2.4 and 0.2.4.1.2 with the
comparison operation now proceeding to compare the fourth segment
index of every path descriptor, but a fourth segment index does not
exist in the case of the first path descriptor.
The next test assumes a situation opposite to the preceding
situation that is the number of segment indices for the second path
descriptor exceeds the number of segment descriptors for the first
path descriptor. In this case the first path descriptor is deemed
to be larger than the second path descriptor. The path comparison
operation stops at this point.
(d) Segment Descriptor. A segment descriptor uniquely identifies a
segment with respect to a path in the context of an electronic
horizon. A segment is identified by the path descriptor of that
path that has the segment to be identified as its last segment. For
example, referring again to FIG. 8, the segment labeled A, can be
identified as 0.2.1.
(e) Node Descriptor. A "node descriptor" uniquely identifies a node
within an electronic horizon. A node descriptor is the path
descriptor of that path that ends in the node to be identified. In
FIG. 8, the node descriptor of the node labeled C is therefore
0.2.2. The node descriptor for the root node of an electronic
horizon has the special value of -1.
(f) Point Descriptor. A point descriptor uniquely identifies any
point within an electronic horizon. A point descriptor consists of
two parts: (1) the segment descriptor of the segment to which the
point belongs and (2) the point index of the point to be
identified. In order to be able to distinguish between point
descriptors and other descriptors, a colon is used to separate the
segment descriptor part of a point descriptor from the point index
itself, e.g., "0.1:2" identifies segment "0.1" and point 2.
(6). Costing functions
(a) Overview
The building of an electronic horizon is that process which
determines which segments (and intersections) are part of an
electronic horizon and which are not. The first segment of an
electronic horizon is the segment on which the vehicle is currently
located. Each time another segment is added to an electronic
horizon the electron horizon calculation process 170(3) determines
whether the exit node of that segment should be expanded further,
i.e., whether any or all of the segments attached to the exit node
of the segment should be also made part of the electronic horizon.
A segment costing function and a node costing function are used for
this purpose.
The costing functions provide how certain factors affect building
of an electronic horizon. The costing functions allow a driver
assistance application (through a configuration process) to specify
whether certain factors should affect building of the electronic
horizon. The costing functions also allow a driver assistance
application to specify (through the configuration process) to what
extent each of these factors should affect building of the
electronic horizon. The following list includes the factors that
can be taken into account by the costing functions.
(1) current vehicle speed;
(2) travel time of the vehicle from the current vehicle
location;
(3) driving distance from the current vehicle location;
(4) inclusion of inaccessible segments;
(5) inclusion of circular paths (e.g., a path having the same
segment entered more than once);
(6) inclusion of U-turns;
(7) inclusion of node costs (e.g., the cost of turns at
intersection); and
(8) inclusion of estimated segment travel costs.
The above list is not exclusive and there may be other factors that
can be considered by the costing functions.
Using these factors, the costing function determines the extents of
an electronic horizon. For example, the extents of the electronic
horizon can include all segments within an absolute distance, all
segments that are reachable at a current speed of the vehicle
within the next n seconds, all segments that are reachable within
the next n seconds while traveling at the legal speed limits of the
corresponding segments, etc. These factors can be combined in
various ways. For example, the extents of an electronic horizon can
include a minimum absolute distance combined with a distance which
is a function of the vehicle speed and time.
(b) The process of computing cost values
The process of building an electronic horizon uses two threshold
cost values. The first threshold cost value is referred to as the
"building threshold cost" and the second threshold cost is referred
to as the "minimum path cost."
The process of computing cost during the building process of an
electronic horizon operates recursively. First, some cost (through
the segment cost function) is associated with the "travel cost" of
the vehicle from the vehicle position (on the first segment of the
electronic horizon) to the exit node of the first segment of the
electronic horizon. The building process now continues in the
following recursive fashion:
For any segment attached to the exit node of the current electronic
horizon segment, a node cost is added. This node cost models the
cost associated with turning from the current onto the attached
segment and it is determined by the node cost function. Then, a
segment cost is added which reflects the cost of the vehicle
travelling from the entrance node of a newly attached segment to
its exit node.
At each step, the current cost is compared with a value for the
"building threshold cost" (or "first threshold"). The building
threshold cost is used as a threshold to determine when the process
extending the path from the current vehicle position should be
stopped.
Once the cost of a path reaches or exceeds the building threshold
cost, the building process stops for that path. Then, the same
building process is applied to the next path, and so on until all
the paths leading out from the current vehicle position are
determined and each path has a cost at least as great to the
building cost threshold, if possible. (Note that in some cases, it
may not be possible to extend a path out to the building threshold
cost. For example, if a road ends in a dead end, the path may end
before the building threshold cost is reached.)
Once an electronic horizon is built, the cost associated with each
of the paths in the electronic horizon is at least as large as the
building threshold cost value (if possible).
As the vehicle travels forward and the vehicle position changes,
data indicating the new position are collected by the sensors (120
in FIG. 1). The vehicle positioning tool (150(2)(1) in FIG. 4) uses
these new data to determine a new vehicle position. Data indicating
the new vehicle position are sent from the vehicle positioning tool
150(2)(1) to the data engine (170 in FIG. 5) where the data are
received by the data receiving component 170(1) which in turn
passes the data to the process 170(3) that calculates the
electronic horizon. Then, the electronic horizon calculation
process 170(3) determines whether a new electronic horizon has to
be built as a result of the new vehicle position or whether the
previous electronic horizon can be reused (Step 170(3)(5)). As part
of making this determination, the electronic horizon calculation
component 170(3) adjusts the costs of all the paths in the
electronic horizon program to take into account the data indicating
the new vehicle position. When adjusting the costs of the paths,
the costs of the paths decrease because the vehicle position
advances into the electronic horizon. At this point, the electronic
horizon calculation process 170(3) determines whether any path in
the electronic horizon has a cost less than the minimum path cost
(i.e., the "second threshold"). If all the paths in the electronic
horizon have costs that exceed the minimum path cost, a new
electronic horizon is not built. Instead, a new electronic horizon
is determined using the paths that had been already determined for
the previous (i.e., existing) electronic horizon. When a new
electronic horizon is determined in this manner, the paths (and
costs thereof) are updated to take into account the new vehicle
position. When a new electronic horizon is determined in this
manner, one or more segments of a path, or even an entire path,
from the previous electronic horizon may be eliminated.
As data indicating new vehicle positions are received in the data
engine 170, the calculation component 170(3) determines new
electronic horizons in this manner until any path cost is less than
the minimum path cost. When a new vehicle position causes any path
cost in an electronic horizon to fall below the minimum path cost
threshold, a completely new electronic horizon is built (i.e., all
the paths starting from the current vehicle position are
determined, in the manner described above, so that the cost of each
path is at least the building threshold cost).
Use of two threshold cost values has several advantages. Using two
cost threshold values provides for a safety margin. This safety
margin is configurable by the driver assistance applications 200
that use the electronic horizon. Another advantage of using two
thresholds is that an entirely new electronic horizon does not have
to be computed as frequently, thereby reducing the computational
requirements associated with the building of the electronic
horizon. Another advantage of using two thresholds is that the
memory required to store the data associated with an electronic
horizon may be reduced (as described below in connection with the
data repository 180).
The values of the building threshold cost and the minimum threshold
cost are configurable. In one embodiment, these values are
configured by the driver assistance applications that use the
electronic horizon.
(c) Computation of the path costs when calculating the electronic
horizon
As stated above, when calculating an electronic horizon, the cost
associated with the addition of each node and segment to the
electronic horizon is determined and added to the costs already
accumulated for the path in order to determine whether expansion of
the electronic horizon along that path should stop. When
determining the cost of adding a segment to a path, the electronic
horizon calculation function 170(3) uses a segment cost function
170(3)(2) and when determining the cost of adding a node to a path,
the electronic horizon calculation function 170(3) uses a node cost
function 170(3)(3).
(d) The segment cost function
The segment cost function 170(3)(2) determines the cost associated
with a vehicle travelling from the entrance node to the exit node
of a segment. In the case of the first segment the cost is limited
to the travel cost of the vehicle from the current vehicle location
to the exit node of the first segment.
According to one embodiment, the segment cost function 170(3)(2)
has access to certain information about a segment for which a cost
is computed. The information about the segment is obtained from the
map database 130. The segment cost function 170(3)(2) may use some
of the data, all the data, or none of the data, depending on how
the segment cost function has been configured. According to one
embodiment, the segment cost function 170(3)(2) has access to the
following information about a segment:
(1). the length ("L") of the segment,
(2). an estimated travel cost ("SETC"), and
(3). whether travel along the segment in the current direction is
legal ("TDI").
(The TDI information allows the driver assistance application to
control, through a configuration process, whether one way streets
oriented opposite to the current vehicle travel direction are
included in an electronic horizon.)
With respect to the first segment, the length is the distance from
the current vehicle location to the exit node of the first segment
and the estimated travel cost is the estimated travel cost from the
vehicle location to the exit node of the first segment.
In the segment cost function 170(3)(2), factors are associated with
combinations of these data items. The segment cost function
170(3)(2) is configured by selecting values for each of these
factors. For example, a legal-length cost factor ("FLEN_Illegal")
can be defined and used as a factor of the segment length ("L") and
the legal travel direction ("TDI"). An illegal-length cost factor
("FLEN_Illegal") can be defined and used as a factor of the segment
length ("L") and the legal travel direction ("TDI"). An
estimated-travel cost factor ("FEST_Legal") can be defined and used
as a factor of the travel cost ("SETC") and the legal travel
direction ("TDI"). Likewise, an illegal-direction-estimated-travel
cost factor ("FEST_Illegal") can be defined and used as a factor of
the travel cost ("SETC") and the legal travel direction
("TDI").
By selection of values for each of these factors, the relative
importance of each of the different types of available information
about a segment can be determined with respect to expansion of the
electronic horizon. In this manner, the segment cost function can
be configured. This configuration may be made based on input from a
driver assistance application or alternatively, default
configuration values may be used.
(e) The node cost function
The electronic horizon calculation process 170(3) also includes a
node cost function 170(3)(3). The node cost function 170(3)(3) is
used to compute the cost associated with the addition of a node to
a path when determining an electronic horizon. The node cost
represents the cost associated with the transition (e.g., turn
right, left, or go straight) from one segment to another.
According to one embodiment, the node cost function 170(3)(3) has
access to certain information about a node for which a cost is
computed. The information about the node is obtained from the map
database 130. The node cost function 170(3)(3) may use some of the
data, all the data, or none of the data, depending on how the node
cost function 170(3)(3) has been configured. According to one
embodiment, the node cost function 170(3)(3) has access to the
following information about a node:
(1). whether the turn across the node is legal ("TL"). A turn can
be illegal because a turn restriction is in place (e.g., no left or
right turn) or the successor segment is a one way street which
would be entered the wrong one way.
(2). the turn angle from the entrance segment to the successor
segment ("TA"). The value may be expressed in degrees.
(3). an estimated node cost ("EnodeCost").
(4). a value ("SecondSegment") which indicates whether the second
segment is already part of the current path for which further
expansion is currently being explored.
As with the segment cost function 170(3)(2), factors can be
associated with these data items in the node cost function
170(3)(3). The node cost function 170(3)(3) is configured by
selecting values for each of these factors. For example, a
turn-angle factor ("F_TA_Legal") is applied to the turn angle
between the current segment and the next segment, if the turn is
legal. (A turn is legal if neither turn restrictions nor one way
restrictions prevent a turn from being executed.) This factor can
be used to associate higher costs with sharper turn angles and vice
versa. A node-cost factor ("F_SDAL_ENodeCost_Legal") can be applied
to the node cost ("EnodeCost") from the database 130. A constant
cost ("Cost_UTurn") can be added in the case the turn is legal and
the turn is a U-turn. By choosing an appropriately high value,
U-turns can be completely suppressed. An illegal-turn factor
("F_TA_Illegal") can be applied to the turn angle between the
current and the next segment if the turn is illegal. A
constant-cost factor ("C_IllegalTurn") can be added if the turn is
illegal. A constant-cost factor ("Cost_SecondSegment") can be added
if the next segment is a segment which is already part of the
current path and both segments have the same direction.
Selection of values for each of the factors in the node cost
function provides for assigning the relative importance of each of
the different types of available information about a node with
respect to expansion of the electronic horizon. In this manner, the
node cost function can be configured. This configuration may be
made based on input from a driver assistance application or
alternatively, default configuration values may be used.
(f) Configuration of the costing functions
As stated above, the building threshold cost and the minimum path
cost may be configured using input from one or more of the driver
assistance applications. These thresholds may be fixed values or
may be computed values. For example, according to one embodiment,
the minimum path cost may be made a function of the vehicle speed.
According to this embodiment, the current vehicle speed ("VSP") is
available from the sensors and continuously updated in the data
provided to the electronic horizon calculation process 170(3).
A value for a minimum path cost factor ("SpeedF") is determined by
one of the driver assistance applications. Using this information,
the minimum path cost ("MinCost") is computed as follows:
The building threshold cost value may also be computed. In one
embodiment, the building threshold cost value ("MaxCost") may be
made a function of the minimum path cost according to the following
relationship:
where MaxF is a factor applied to the minimum path cost.
As mentioned above, the costing functions may be configured using
input from the driver assistance application. One way to configure
the costing functions is to ensure that all paths within the
electronic horizon have a certain minimum length, that U-turns are
to be ignored, and that inaccessible segments are not made part of
an electronic horizon. This setup can be achieved as follows:
In the segment cost function, FLEN_Legal is set to 1. This makes
the cost identical to a segment's length (or in the case of the
first segment identical to the distance of the vehicle to the exit
node of the first segment). Also in the segment cost function,
FLEN_Illegal is set to zero to suppress inaccessible segments. Also
in the segment cost function, FEST_Legal and FEST_Illegal are both
set to zero. This way, any estimates of travel times are ignored.
In the node cost function, F_TA_Legal and F_SDAL_ENodeCost_Legal
are set to 0 thereby eliminating any costs for legal turns.
Cost_Uturn is set to 100,000 to eliminate any U-turns. F_TA_Illega
is set to 0 but C_Illegal_Turn is set to 100,000 to eliminate any
inaccessible segments or illegal turns. Cost_SecondSegment is set
to 100,000 to eliminate the same segment being part of any path
twice.
(7). Primary Path
(a) Overview
Some driver assistance applications require the processing of all
possible paths within an electronic horizon (i.e., accessible and
inaccessible paths). However, some driver assistance applications
use a "primary path." A "primary path" is one specific path of the
one or more potential paths within an electronic horizon. The
primary path is the most likely path upon which the vehicle is
expected to travel. The data horizon program 110 includes a feature
by which a primary path can be determined and identified to a
driver assistance application.
There are two aspects to the computation of the primary path. A
first aspect is an estimation of the most likely driving path based
on the local road geometry. A second aspect is the use of route
information, if available. These aspects are discussed below.
(b) Most Likely Path
The data engine 170 of the data horizon program 110 includes a
primary path function 170(6). Included in the primary path function
is a function 170(6)(1) that calculates a local-road-network-based
most likely path ("LRNBMLP"). The function 170(6)(1) attempts to
estimate how the vehicle will continue to travel within the current
electronic horizon taking into account only the local road network.
The function 170(6)(1) computes a single path as the LRNBMLP. The
function 170(6)(1) computes the LRNBMLP as follows. The function
170(6)(1) includes the first electronic horizon segment in the
LRNBMLP. Then, the following steps for the selection of the next
segment are repeatedly executed by the function 170(6)(1) until a
leaf node of the electronic horizon is found.
If only one accessible segment is attached to a node, that segment
is chosen.
If more than one accessible segment is attached to a node, then
from among all accessible segments the segment with the highest
functional class is chosen. If two or more accessible segments have
the same functional class which is higher than the functional class
of each of the other segments, the segment with the highest
functional class with the smallest turn angle is chosen. If there
are two segments with the highest functional class and the same
turn angle (e.g., one being a left and the other being a right
turn), the right turn is chosen over the left turn.
A driver assistance application may chose to have a LRNBMLP
determined in this manner. Alternatively, the driver assistance
application may chose not to have the LRNBMLP determined in this
manner.
(c) Route-based path
As mentioned above, another aspect of determining a primary path of
a vehicle is to use route information. Some vehicles include
hardware and software that can calculate a route to desired
destination. As mentioned above in connection with FIG. 4, in a
present embodiment, a route calculation tool 150(2)(3) is included
among the navigation applications 150(2). The route calculation
tool 150(2)(3) can be used to calculate a route to a desired
destination. In one embodiment, the route calculation tool
150(2)(3) provides an output in the form a data route ("R"). The
data route is a list of consecutive and directed segments
describing a legal way for a vehicle to drive from the first to the
last segment of the route. A "route sub-path" of a route within
some given electronic horizon is that path within the electronic
horizon that matches some (or all) segments of a given route. Given
a route, it is possible that the route is not contained (at least
partially) within the electronic horizon. In this case, the route
sub-path is undefined (and therefore identified by the path
descriptor of -1).
The primary path function 170(6) includes a function 170(6)(2) that
attempts to calculate a route-based path. A route-based path is
that part of a calculated route which is located within an
electronic horizon. Inputs to the function 170(6)(2) include data
indicating the route R and data ("E") indicating the calculated
electronic horizon. As a first step, the function 170(6)(2)
determines whether a route-based path can be defined for the
electronic horizon. To perform this step, the function 170(6)(2)
attempts to locate the first segment of the electronic horizon in
the calculated route R. If the first segment of the electronic
horizon cannot be found in the calculated route R, the computation
stops and the route-based path is undefined (i.e., there is no
route-based path). However, if the first segment of the electronic
horizon matches one of the segments in the calculated route R, the
route-based path is defined. (Note that in order for the first
segment of the electronic horizon to match one of the segments in
the calculated route, the function 170(6)(2) requires that the
direction of travel along the segment in both the electronic
horizon and the calculated route be the same.) After the first
segment of the electronic horizon is found in the calculated route,
the function 170(6)(2) continues to attempt to match segments from
the paths in the electronic horizon E with segments from the
calculated route R. As with the first segment, the function
170(6)(2) requires that the direction of travel on the matching
segments be the same. This matching process continues until no more
segments from the paths of the electronic horizon can be found
among the segments of the route. Matches are no longer found
because a segment from the route for which a match is sought in E
is not contained in E (i.e., because the electronic horizon E does
not extend beyond some node) or the last segment of the route R was
reached and therefore no additional segments of R can be matched in
E.
(d) Computing the primary path
The primary path computation function 170(6) computes a primary
path using the outputs from the most likely path function 170(6)(1)
and the route-based function 170(6)(2). If a route R has been
defined and the route-based function 170(6)(2) was able to
determine a route-based path based on R, then that route-based path
of R is selected as the primary path. However, if either a route
has not been defined or it was not possible to compute a
route-based path, the local road network most likely path (LRNBMLP)
is used. An advantage of this method is to assume that the driver
will follow a calculated route, if he/she has entered route
information. However, if no route information is available, the
local road network most likely path is the best estimate that can
be provided.
(8). Determining the contents of the newly built electronic
horizon
Reference is made again to FIG. 5. When the calculation process
170(3) has built a new electronic horizon (as opposed to
determining a new electronic horizon by adjusting the vehicle
position and path costs from the previous electronic horizon), the
contents for the new electronic horizon data structure are
obtained. The data engine 170 includes a component process 170(4)
that performs this function. The process 170(4) receives data from
the electronic horizon calculation process 170(3) that indicates
the paths (and consequently which segments and nodes) are to be
represented in the electronic horizon data structure. Upon
receiving this data, the electronic horizon content formation
process 170(4) obtains from the map database 130 the necessary data
for formation of the electronic horizon data structure. The data
structure formed by the electronic horizon content formation
process 170(4) contains the relevant data about the roads and
intersections in the electronic horizon. This data structure forms
the output 171 of the data engine 170.
The types of data that the electronic horizon content formation
process 170(4) obtains from the map database 130 are determined by
a configuration process. This configuration process may be
performed during a manufacturing stage of the advanced driver
assistance systems or during an initialization or setup process of
the advanced driver assistance systems. In one embodiment, the
configuration controller 165 receives data from one or more driver
assistance applications 200 that indicate the types of data that
should be included in the electronic horizon. In turn, the
configuration controller 165 provides data to the process 170(4) to
indicate the types of data attributes associated with segments and
nodes should be obtained for inclusion in the electronic horizon
data structure. Based on these inputs, the content formation
process 170(4) obtains the necessary data from the map database 130
to include in an electronic horizon data structure whenever a new
electronic horizon is built.
When a newly built electronic horizon data structure have been
obtained and stored in the appropriate structure, the contents of
the structure are output from the data engine 170 to the data
repository 180. The data engine 170 includes a component process
170(8) that provides this output 171.
As mentioned above, under some circumstances (e.g., an off-road
condition), an electronic horizon cannot be calculated. If an
electronic horizon cannot be calculated, the process 170(4) does
not obtain any data for an electronic horizon data structure from
the map database 130. Under these circumstances, the content
formation process 170(4) provides no output or alternatively the
content formation process 170(4) provides an empty electronic
horizon, i.e., indicating that no electronic horizon has been
determined for the vehicle position.
There is another occasion when an empty electronic horizon is
provided. It is possible that the vehicle positioning tool
150(2)(1) reports the vehicle to be traveling against the legal
driving direction of a one way street. In this case, the electronic
horizon calculation process 170(3) returns the appropriate state
information with the electronic horizon in essence being empty.
If the calculation process 170(3) has been configured to provide a
primary path (instead of an entire electronic horizon), the
electronic horizon content formation process 170(4) obtains the
data from the map database 130 needed for an electronic horizon
data structure that includes only the primary path. (Alternatively,
an electronic horizon including all the paths is provided along
with data separately indicating the primary path.) If the
calculation process 170(3) has been configured to provide a primary
path and a primary path cannot be determined, the content formation
process 170(4) provides an output indicating that no primary path
has been determined for the vehicle position.
In the embodiment shown in FIG. 5, the process 170(4) of obtaining
data for the electronic horizon is shown as separate from the
process 170(3) of determining the electronic horizon. In
alternative embodiments, these processes may be combined so that
the data contained in the electronic horizon is obtained and the
electronic horizon is built as the paths that make up the
electronic horizon are determined.
(8). Providing the electronic horizon
As mentioned above, according to a present embodiment, a new
electronic horizon is not necessarily built each time a new vehicle
position is obtained. Instead, the previous electronic horizon can
be reused if all the path costs of the previous electronic horizon
still exceed the minimum threshold cost after adjustment for a new
vehicle position. Under these circumstances, the data engine 170
provides an output 172 indicating a new electronic horizon for the
new vehicle position that uses the paths determined for a previous
electronic horizon. The data engine 170 includes an electronic
horizon output process 170(7) that performs this function. The
electronic horizon output process 170(7) provides this output 172
to the data repository 180, as explained in more detail below.
According to one embodiment, the electronic horizon output process
170(7) provides an output for each receipt of data indicating a new
vehicle position. According to a present embodiment, the component
170(7) that provides the output 172 defining an electronic horizon
is separate from the component 170(8) that provides the contents of
an electronic horizon. The output 171 of the electronic horizon
content output process 170(8) includes all the necessary data
attributes associated with all segments and nodes in all the paths
forming an electronic horizon. The output 172 of the electronic
horizon output process 170(7) includes only a reference to one of
the outputs 171 that contains the data contents of an electronic
horizon and an indication of the vehicle position relative to the
referenced data contents.
F. The data repository 180
As stated above in connection with FIG. 1, the data repository 180
is the component of the data horizon program 110 that contains the
latest data readings. An embodiment of the data repository
component 180 is shown in FIGS. 10-12. As shown in FIG. 10, the
data repository 180 contains three different types of data. First,
the data repository 180 holds data 180(1) representing the
electronic horizon that had been determined by the data engine 170.
According to one embodiment, the data 180(1) includes the attribute
information about the segments and nodes in the electronic horizon.
The attribute information about the segments and nodes in the
electronic horizon may include some or all the attributes
identified in FIGS. 3A and 3B. Second, the data repository 180
holds data 180(2) representing the vehicle position. The data
180(2) representing the vehicle position is that data determined by
the vehicle positioning tool (150(2)(1) in FIG. 4). The data
repository 180 may obtain the data 180(2) representing the vehicle
position directly from the vehicle positioning tool 150(2)(1) or
the data 180(2) representing the vehicle position may be obtained
from the data engine 170. Third, the data repository 180 holds
sensor data 180(3). The sensor data 180(3) may be raw sensor data
obtained directly from the sensors (120 in FIG. 1) or alternatively
the sensor data 180(3) may be obtained from the data engine
170.
Referring to FIG. 11, with respect to the electronic horizon data
180(1), the data repository 180 holds at least the set of data
representing the most recent electronic horizon that had been
determined by the data engine 170. In one embodiment, the data
repository 180 holds several sets of data representing several
electronic horizons. These several sets of data held in the data
repository 180 are those sets created most recently by the data
engine 170. For example, the data repository 180 may hold the ten
most recent sets of data representing the ten most recent
electronic horizons that had been determined by the data engine 170
although a number greater or lesser than ten may also be suitable.
The number of sets of data retained by the data repository 180 may
be configured using input from the driver assistance applications
200 via the configuration controller (165 in FIG. 1). Each set of
data in the data repository 180 is assigned an identification
number or code by which it can be identified.
According to an embodiment shown in FIG. 11, the data repository
180 does not necessarily hold complete sets of data for each
electronic horizon retained therein. Instead, the data repository
180 implements a handle-container mechanism. This handle-container
mechanism is similar to mechanisms used in object oriented
programming to handle large objects. Using the handle-container
mechanism reduces the storage and handling requirements for
multiple sets of data representing corresponding multiple
electronic horizons.
Use of a handle-container mechanism for storage of electronic
horizons in the data repository 180 is facilitated by the manner in
which electronic horizons are calculated by the data engine 170. As
mentioned above, according to one embodiment, a new electronic
horizon is not necessarily built each time data indicating a new
vehicle position is obtained. Instead, a new electronic horizon is
built only when a path from the previous electronic horizon falls
below a minimum path threshold after taking into account a new
vehicle position.
According to the embodiment shown in FIG. 11, a class
ElectronicHorizonData is defined and a class ElectronicHorizon is
defined. The objects 181 in the ElectronicHorizonData class contain
all the information (i.e., data attributes) needed to represent an
electronic horizon. Additionally, each ElectronicHorizonData object
181 contains a reference count. The reference count indicates how
many other objects are using the ElectronicHorizonData object
181.
Each object 182 in the ElectronicHorizon class contains three
pieces of information: a pointer, a delta distance, and a handle
(i.e., ID). The pointer points to the applicable
ElectronicHorizonData object 181. The delta distance in an
ElectronicHorizon object 182 is a value that indicates the
difference in the vehicle position of the ElectronicHorizon object
182 relative to the vehicle position in the referenced
ElectronicHorizonData object 181. (As long as the vehicle remains
on the same segment and has moved such that the most recently used
electronic horizon data can be reused, no new electronic horizon
data is computed.)
Use of the handle-container mechanism for storage and use of
electronic horizons affords several advantages. Electronic horizons
would take up a lot of memory if they were stored as ordinary class
objects. However, in the embodiment of FIG. 11, the
ElectronicHorizon object 182 contains only three items of
information and accordingly may be relatively small compared to the
ElectronicHorizonData object 181. Copying an ElectronicHorizon
object 182 implies copying the data contained in the
ElectronicHorizon object 182, but as far as the associated
electronic horizon data is concerned, only a pointer to the
respective ElectronicHorizonData object 181 is copied. When the
ElectronicHorizon object 182 is copied, the reference count in the
applicable ElectronicHorizonData object 181 is incremented
indicating that the ElectronicHorizon object 182 is using the data.
An ElectronicHorizonData object 181 is deleted when all
ElectronicHorizon objects 181 referring to it cease to exist.
Reference is made to FIG. 12. As stated above, the data repository
180 also contains vehicle position data 180(2). The vehicle
position data 180(2) contained in the data repository 180 includes
data indicating the most recent one or more vehicle positions that
had been determined by the vehicle positioning tool (150(2)(1) in
FIG. 4). The number of vehicle positions included in the vehicle
position data 180(2) retained by the data repository 180 may be
configured. In one embodiment, the number of vehicle positions
represented by the vehicle position data 180(2) contained in the
data repository 180 corresponds to the number of electronic
horizons included in the electronic horizon data 180(1).
Alternatively, the number of vehicle positions represented in the
vehicle position data 180(2) contained in the data repository 180
may be greater than the number of electronic horizons included in
the electronic horizon data 180(1). The vehicle position data
180(2) may be retained separately from the electronic horizon data
180(1) or alternatively the vehicle position data 180(2) may be
included with the electronic horizon data 180(1). As shown in FIG.
12, each set of vehicle position data 180(2) may be assigned an
identification number or code by which it can be identified.
Also as stated above, the data repository 180 contains sensor data
180(3). The sensor data 180(32) contained in the data repository
180 includes the most recent sensor readings from the sensors 120
(in FIG. 1). The number of sensor readings included in the data
repository 180 may be configured. In one embodiment, the number of
sensor readings contained in the data repository 180 corresponds to
the number of electronic horizons included in the electronic
horizon data 180(1) or the number of vehicle positions included in
the vehicle position data 180(2). Alternatively, the number of
sensor readings contained in the data repository 180 may be a
different number. As shown in FIG. 12, each set of sensor data
180(3) may be assigned an identification number or code by which it
can be identified.
In addition to the electronic horizon data 180(1), the vehicle
position data 180(2) and the sensor data (3), the data repository
180 may also contain other kinds of data.
G. The data distributor 190
FIG. 13 shows the components of the data distributor 190. The data
distributor 190 is that component of the data horizon program 110
that initiates the sending of data from the data repository 180 to
the driver assistance applications 200 that use the data. In order
to reduce processing requirements, the data distributor 190
includes a component 190(1) that sends messages 191 indicating the
availability of new data. These messages are sent to over a vehicle
data bus 194 to each driver assistance process 200 that uses data
stored in the data repository 180. In an embodiment in which there
are several driver assistance processes 200 that use data stored in
the data repository 180, the messages 191 from the data distributor
190 are broadcast over the data bus 194 to each process 200 that
uses the data. Each driver assistance process 200 that uses data
stored in the data repository 180 registers with the data
distributor 190 to receive the messages about the availability of
new data.
With respect to electronic horizon data (180(1) in FIG. 10), the
data distributor 190 broadcasts messages about the availability of
new data once each cyclic execution of the data engine 170. With
respect to the vehicle position data 180(2) and the sensor data
180(3), the data distributor 190 broadcasts messages about the
availability of new data when such new data becomes available.
Each message 191 identifies the availability of new data by an ID
(or pointer). For example, with respect to the electronic horizon
data 180(1), the message 191 sent by the data distributor 190 to
the driver assistance applications 200 that use the data includes
the ID associated with the electronic horizon data 180(1) in the
data repository 180. Each message 191 may also indicate the type of
new data which is available, e.g., electronic horizon, vehicle
position, or sensor.
(The data distributor 190 also includes a registration component
190(2). The registration component 190(2) is used in conjunction
with corresponding registration components 302 in the listeners
300, as explained in more detail below.)
H. The data listener 300
In the embodiment shown in FIG. 1, each of the driver assistance
applications 200 that use the data collected by the data horizon
program 110 uses a data listener 300. A data listener 300 is a set
of functions that is associated with an driver assistance
application 200 that uses the data collected by the data horizon
program 110. A data listener 300 provides a means by which a driver
assistance application 200 interfaces with the data horizon program
110. The data listener 300 includes processes by which each driver
assistance application 200 that uses data stored by the data
horizon program 110 can obtain the data it requires.
FIG. 14 shows components of a data listener 300(n). The data
listener 300(n) is shown associated with a driver's assistance
application 200(n). As shown in FIG. 14, the data listener 300(n)
includes a registration component 302. The registration component
302 registers the particular listener 300(n) with the data horizon
program 110. Specifically, the registration component 302 registers
with the registration component 190(2) of the data distributor 190.
As part of the registration process, the registration component 302
transmits a message to the data distributor 190 indicating that the
listener (of which the component 302 is a part) is to be notified
about the availability of new data. As part of the registration
process, the registration component 302 also identifies to the
registration component 190(2) of the data distributor 190 the type
of data that about which the listener 300(n) is to be notified
(e.g., electronic horizon data, vehicle position data, or sensor
data.) In the embodiment of FIG. 14, the listener 300(n) is used
for notification of electronic horizon data. Once the listener
300(n) is registered with the data distributor 190, the listener
300(n) will continue to be sent notifications from the data
distributor 190 about the availability of new data of the type
specified during registration as the new data is deposited in the
data repository 180. The registration process may be performed
once, e.g., when the driver assistance application 200 is
initialized. The registration process may be performed subsequent
times.
As stated above, after the listener 300(n) is registered with the
data distributor 190, the listener 300(n) is regularly sent
notifications 191 about the availability of new data. The data
listener 300(n) includes a component 304 that receives these
notifications 191. As mentioned above, each notification 191
includes an identification (i.e., ID) of a set of new data stored
in the data repository 180. The data listener 300(n) includes a
component 306 that stores each identification in a queue 310. The
queue 310 is included as part of the data listener 300(n). The
identifications stored in the queue 310 include at least those from
the latest notifications received from the data distributor 190.
The queue 310 may include identifications from several of the most
recent notifications received from the data distributor 190. The
size of the queue is configurable.
When the application 200(n) is ready to receive new data, the data
listener 300(n) obtains the new data for the application 200(n).
The data listener 300(n) includes a component 312 that obtains an
identification from the queue 310. The component 312 may obtain the
most recent identification added to the queue 310 or alternatively,
the component 312 may obtain any other identification the queue
310. Upon obtaining an identification from the queue 310, a process
314 in the data listener 300(n) uses the identification to obtain
the associated data from the data repository 180. Upon receiving
the data from the data repository 190, a process 316 in the data
listener 300(n) provides the data to the driver assistance
application 200(n).
A driver assistance application 200 may use more than one of the
different types of data stored in the data repository 180. If a
driver assistance application uses more than one of the different
types of data stored in the data repository 180, the driver
assistance application is associated with more than one data
listener. According to one embodiment, a separate data listener 300
is used by a driver assistance application for each of the
different types of data that the driver assistance application
uses. For example, if a driver assistance application 200 uses both
electronic horizon data and sensor data, the driver assistance
application 200 is associated with two separate data listeners
300--one for the electronic horizon data and the other for the
sensor data. Each of the data listeners associated with a single
driver assistance application receives messages from the data
distributor 190 of the data horizon program 110 about the
availability of new data of the type associated with the listener.
Each of the data listeners maintains a separate queue of ID's by
which the respective types of data can be obtained from the data
repository 180. FIG. 15 shows an embodiment of a driver assistance
application 200(k) associated with three separate listeners
300(k)(1), 300(k)(2), and 300(k)(3), for obtaining three different
kinds of data.
I. Alternative embodiment for listener
In an embodiment described above, a data listener 300 was disclosed
as a separate object from the driver assistance application 200
associated therewith that uses the data for which the listener was
receiving notifications. According to an alternative embodiment,
the listener function can be incorporated into the same object that
processes the data for which the listener receives notifications.
According to this alternative, an object (or application) that
receives notifications about new data (from the data distributor)
also directly processes the data. An application that both receives
notifications about data and processes the data about which it
receives notifications can implement these two functios as separate
threads.
As described above in connection with the embodiment in which the
listener process is implemented as a separate application or
object, the event notification mechanism used in the listener
requires that a notification call by the data horizon program
return quickly. A notification call should consume minimual
processing time and only signal the availabilty of data or start a
thread that will get the data. In an embodiment in which the
listener function is implemented as a separate thread in the same
application or object that also implements the processing of the
data, the event notification mechanism should also return quickly.
In addition, in an embodiment in which the listener function is
implemented as a separate thread in the same application or object
that also processes the data, a means is used to start or stop the
thread that performs the listener function. This can be performed
by the data horizon program. Specifically, the data engine 170 can
invoke the thread that listens for the event notification within
the application or object that uses the data. A process in an
application that uses the data can be registered with the data
engine 170 in a similar fashion as described above in connection
with a listener. Once the listener thread has been registered, the
data engine starts (or stops, suspends or resumes) this thread
whenever the data engine is started (stopped, suspended or
resumed).
J. The monitoring program 160
Referring again to FIG. 1, the monitoring program 160 is a part of
the data architecture 100. The monitoring program 160 allows for
viewing the execution of the functions of the data horizon program
110. Some of the features of the monitoring program 160 may be used
in a testing and configuration environment. Other features of the
monitoring program 160 may be used during ordinary use by an end
user of the motor vehicle 108 in which the map data architecture
100 is installed. In one alternative, the monitoring program 160 is
used only in a testing and configuration environment and not in a
run time environment (e.g., during ordinary operation of the
vehicle by an end user).
In a testing and configuration environment, an output of the
monitoring program 160 may be provided to a display monitor 160(1)
on which various aspects of the execution of the functions of the
data horizon program 110 can be viewed. For example, the monitoring
program 160 may present a continuous image of the position of the
moving vehicle on a map on the display monitor 160(1). The display
monitor 160(1) may also show an area around the current location of
the vehicle. Those road segments that are parts of paths in the
electronic horizon may be highlighted on the display monitor
160(1). In addition, the monitoring program 160 may show the
current vehicle position, including a spot, heading, and speed on a
map image on the display 160(1). If the vehicle 108 is following a
route calculated by the route calculation tool (150(2)(3) in FIG.
4), the calculated route may be highlighted on the map image on the
display 160(1). In addition, the monitoring program 160 may display
the attributes of the road segments and intersections around the
vehicle. These attributes include the attributes shown in FIGS. 3A
and 3B. Attributes associated with the electronic horizon may also
displayed. The monitoring program 160 adjusts the boundaries of the
image of the map on the display monitor based on the current
vehicle movement.
K. The configuration program 165
Referring again to FIG. 1, the configuration controller program 165
is a part of the data architecture 100. The configuration
controller program 165 allows for configuring of the functions of
the data horizon program 110. The configuration controller program
165 provides for setting the parameters, defaults, etc., that
control the operation of data architecture 100, including the data
horizon program 110. For example the configuration program 165
provides for determining the size of the electronic horizon in
front of the vehicle for which data readings will be
determined.
The configuration program 165 may provide for setting parameters
during installation (or manufacture) of the driver assistance
system equipment in the vehicle. The configuration program 165 may
also provide for setting parameters when new equipment is
installed, e.g., new sensors, new hardware, more memory. The
configuration program 165 may also provide for setting new
parameters when new data is installed, e.g., when the database 130
is updated.
The configuration program 165 may also be used at initialization or
during operation of the vehicle in order to change the operating
characteristics of the data horizon program 110. The configuration
program 165 may receive inputs automatically from the driver
assistance applications 200. The driver assistance applications 200
provide outputs indicating the types of data that they need. The
driver assistance applications 200 may also provide outputs
indicating the extents needed for the electronic horizon. The
extent of the electronic horizon may be specified in distance
(e.g., meters) or time (e.g., segments onto which the vehicle can
travel within the next 10 seconds).
The configuration program 165 can be used to register a data
listener 300 to receive a continuous broadcast of the latest data
values from the data distributor 190.
The configuration program 165 can also be used to interface the
data listener 300 to an in-vehicle data bus architecture for
transfer of data readings to the vehicle's advanced driver
assistance applications 200 running on the bus.
L. Using the advanced driver assistance system map data
architecture
(1) Overview
Advanced driver assistance systems provide ways to improve the
safety, comfort, efficiency, and overall satisfaction of driving.
These systems require information about the road network around the
vehicle. Some of this information can be obtained by sensors.
However, sensors do not reliably obtain all the types of
information needed by some of these systems. Accordingly, use of a
map database in addition to, or as a substitute for, sensors can
make advanced driver assistance systems operate better and more
reliably.
Embodiments of the disclosed advanced driver assistance systems map
data architecture (100 in FIG. 1) provide a means by which one or
more advanced driver assistance system applications 200 can use map
data in support of the function(s) provided thereby. The advanced
driver assistance systems map data architecture provides advanced
driver assistance system applications with access to data about
road geometry and other attributes within the vicinity of the
vehicle. For example, the advanced driver assistance systems map
data architecture provides access to data representing any location
along the road network near the vehicle that can be reached within
10 seconds of driving time. This portion of the road network
corresponds to the electronic horizon. The electronic horizon is
re-calculated regularly over time and/or as the vehicle moves along
the road network. Once an electronic horizon has been calculated,
the advanced driver assistance system application can use the data
about the vehicle paths in the electronic horizon.
Referring to FIG. 16, an advanced driver assistance application 200
can access the data represented by an electronic horizon with an
electronic horizon handle (i.e., the ID of the electronic horizon
object 182 in the data repository 180 in FIG. 11). The advanced
driver assistance application 200 relies on the listener (300 in
FIG. 14) to obtain the ID of the latest electronic horizon (182 in
FIG. 11) from the data distributor 190. With the ID of the
electronic horizon object 182, any or all of the data in the
electronic horizon data object (181 in FIG. 11) can be obtained.
The electronic horizon data object 181 identifies all the possible
vehicle paths (or the primary path) out to the extent of the
electronic horizon. The electronic horizon data object 181 also
identifies the segments and nodes in each path (i.e., using the
segment descriptors and node descriptors, described above).
According to a present embodiment, advanced driver assistance
applications may also obtain sensor data and vehicle position
data.
(2) Iterators
With respect to data contained in an electronic horizon, a driver
assistance application 200 can use one of the iterators to obtain
the data contained in an electronic horizon in an organized manner.
An iterator is a programming construct that allows the successive
retrieval of items from a collection of items. The iterators allow
an advanced driver assistance application to traverse the
electronic horizon for retrieval of path descriptors, electronic
horizon segments, data about points along segments, etc. Included
among the iterators that are available for use by the driver
assistance applications are a path iterator 402, a segment iterator
404 and a segment point iterator 406. To use any of the iterators,
the advanced driver assistance applications initialize the iterator
with the appropriate electronic horizon ID and/or other appropriate
information.
(a) Path iterator
The path iterator 402 is an iterator that generates all paths of an
electronic horizon, one path at a time. The path iterators allows
the generation of all paths or only of those paths which are
accessible.
(b) Segment iterator
The segment iterator 404 returns a list of electronic horizon
segments. Given a node, the segment iterator 404 first returns the
entrance segment of that node (in the context of a path in the
electronic horizon) and then all exit segments of the node (in
clockwise orientation).
(c) Segment point iterator
The segment point iterator 406 is an iterator that returns segment
points. A segment point iterator 406 can be initialized with a
segment of an electronic horizon or with a path of an electronic
horizon. When initialized with a segment of an electronic horizon,
the segment point iterator 406 returns all points of the segment
starting with the entrance node of the segment. When initialized
with a path of an electronic horizon, the segment point iterator
406 returns the first point after the current vehicle position and
then all the points along all the segments that form the path in
the order in which they occur in the path. Note that for an
intermediate node of a path, the segment point iterator 406 returns
first the exit node of the incoming segment and then the entrance
node of the outgoing segment.
(3) Determining the accuracy of data
In some of the embodiments of the map database (130 in FIGS. 1, 3A
and 3B) some roads are represented by higher accuracy data than
other roads. Some advanced driver assistance systems 200 may
require that the vehicle be located on a road represented by the
higher accuracy data. Alternatively, some advanced driver
assistance systems 200 may require that all the roads located
around the vehicle (e.g., in the electronic horizon) be represented
by the higher accuracy data. Thus, the architecture 100 provides a
means by which the driver assistance applications 200 can determine
whether the vehicle is located on a road represented by higher
accuracy data or whether all the road segments located within the
electronic horizon are represented by higher accuracy data. If the
higher accuracy data is located in a supplementary database, such
as the database 130(2) in FIG. 1, the determination whether the
data is higher accuracy data can be made identifying the source of
the data (e.g., the supplementary database 130(2) or the primary
database 130(1)). In a single database embodiment having both
higher accuracy data and lower accuracy data, the determination
whether the data is higher accuracy can be made by reference to an
appropriate data attribute (such as the accuracy level attribute,
described above). In some embodiments, the data horizon program 110
can be configured not to provide an electronic horizon unless all
road segments in all the paths of the electronic horizon are
represented by higher accuracy data.
M. Implementation
The advanced driver assistance systems data interface architecture
includes software and hardware components that run on a suitable
computing platform. In a prototype system, the advanced driver
assistance systems data interface architecture runs in a Microsoft
Windows or Microsoft NT environment including a networked personal
computer (Pentium II or higher). Alternative platforms are also
suitable.
In a prototype environment, data is passed from the sensors 120 to
the connected personal computer via a serial connection
(RS-232).
III. ALTERNATIVE EMBODIMENTS
A. In-vehicle bus architecture alternative
An alternative embodiment of the driver assistance systems map data
architecture 600 is shown in FIG. 17. According to this
alternative, driver assistance applications 602 run on dedicated
micro-controllers connected to an in-vehicle data bus 610. In this
embodiment, the in-vehicle data bus 610 is a CAN bus although in
alternative embodiment, the in-vehicle bus can be any other kind of
bus. In the embodiment of FIG. 17, a data listener 620 (which may
be similar or identical to the data listeners 300, described above)
is adapted to interface to the in-vehicle data bus 610 and
communicate data readings using the standard methods and protocols
for that bus.
B. Electronic horizon combined with sensor data
In the embodiment of the data horizon program described above, an
electronic horizon data object was formed that included data
representing the paths that the vehicle can follow out to the
extents of the electronic horizon. The data representing the paths
included data representing road attributes, road geometry, and road
objects. In the embodiment described above, the data representing
the paths was either obtained from the map database 130 or derived
from data in the map database (e.g., curvature). According to an
alternative embodiment, the electronic horizon data also includes
dynamic data. Dynamic data includes data from the sensors, derived
from the sensor data, or derived from a combination of sensor data
and data from the map database. According to this embodiment,
sensor data can be associated with one or more of the paths in the
electronic horizon. As an example, if a radar system sensor in the
vehicle detects an object located 100 meters ahead of the vehicle,
data indicating this detected object is included in the electronic
horizon. If an electronic horizon path corresponds to the location
of the detected object, data indicating the detected object may be
associated with the path at the corresponding location (e.g., at a
point of the segment in the path).
According to another aspect, if a feature represented by data in
the map database should be detectable by one or more sensors in the
vehicle, a routine in the data horizon program attempts to match
the represented feature to an object detected by the sensors. For
example, assume the electronic horizon includes data from the map
database indicating the presence of an overpass located 80 meters
ahead of the vehicle and further assume that a radar system sensor
in the vehicle detects an object located 82 meters ahead of the
vehicle extending across the road. According to this alternative, a
routine in the data horizon program relates the data from the map
database indicating the presence of the overpass and the data from
the radar sensor indicating the presence of an object extending
across the road. According to a further aspect of this alternative,
a routine in the data horizon program may indicate a difference
(e.g. a .DELTA.) between the location of the overpass as indicated
by the data from the map database and the location of the object
extending across the road as indicated by radar sensor.
C. Other alternatives
In the embodiment of the data repository described above, a
handle-container mechanism was described that facilitated storage
and use of the electronic horizon data. In alternative embodiments,
each set of data that represents a separate electronic horizon may
be retained as a full set of data (i.e., all the attributes for
each path).
As mentioned above, the data engine 170 (in FIG. 5) may be
configured to determined a primary path. If a primary path has been
determined by the data engine 170, the electronic horizon data
(180(1) in FIG. 10) contained in the data repository 180 may
include only the primary path data. Alternatively, the data
repository 180 may also contain both the primary path data as well
as data representing the entire electronic horizon data.
As mentioned above, in one embodiment, a separate data listener is
used by a driver assistance application for each of the different
types of data that the driver assistance application uses.
According to an alternative embodiment, a single data listener may
be used for more than one type of data. According to this
alternative, a single data listener receives notifications about
more than one type of data and responds with requests for more than
one type of data. For example, according to this alternative
embodiment, if a driver assistance application uses both electronic
horizon data and sensor data, a single data listener can be
associated with the driver assistance application and be used to
receive notifications about both the electronic horizon data and
the sensor data.
In an embodiment described above, a listener receives a
notification about the availability of new data in the data
repository and then requests the new data be sent to it. According
to an alternative embodiment, when a listener receives a
notification about the availability of new data, it can request
that the new data be sent by broadcast, multicast, or other means,
to several applications and/or listeners.
According to a further alternative embodiment, a data listener
registers with the data distributor and thereafter automatically
receives the data in the electronic horizon when it becomes
available. According to this alternative, the data listener does
not first receive a notification of the availability of new data
and request the new data upon receipt of the notification.
According to this alternative embodiment, the data listener can
receive the new data by point-to-point transmission, broadcast,
multicast, or other means.
IV. ADVANTAGES
The embodiments of the advanced driver assistance system map data
architecture (in FIGS. 1 and 17) provide a means by which one or
more advanced driver assistance systems can utilize map data to
support the function(s) provided thereby. Use of map data by
advanced driver assistance systems can enhance the functions
provided by such systems. The architecture disclosed herein affords
a means by which more than one driver assistance application can
use the same map data. The architecture disclosed herein also
affords a means by which different driver assistance applications
can obtain different kinds of map data. In addition, the
architecture disclosed herein also affords a means by which
different driver assistance applications can obtain map data at
different rates.
Embodiments of the map data architecture disclosed herein provide
additional advantages. Driver assistance application software is
maintained separate from the data horizon program, thereby
providing versatility, compatibility, and reliability. Moreover,
because the data horizon program implements an easy to use
interface, it is relatively easy for differnet kinds of driver
assistance applications to use the data horizon program.
It is intended that the foregoing detailed description be regarded
as illustrative rather than limiting and that it is understood that
the following claims including all equivalents are intended to
define the scope of the invention.
* * * * *